The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 7 ~
Paper Code : M2 – 05
FINTECH AND DIGITAL BANKING: PERCEPTION AND USAGE IN MAURITIUS,
A LOGISTIC REGRESSION APPROACH
Bhavish Jugurnath
Roucheet Bissessur
Preesha Ramtohul
Ramen Mootooganagen
Faculty of Law and Management, University of Mauritius, Mauritius
Abstract
International trends have manifested mobile banking as a success in terms of promoting financial inclusion.
However, mobile banking is witnessed as escalating from its infancy stage to become a financial revolution
with larger commercial banks expanding its performance from being query based to being actionable
based.
The present study uses the constructs of empirical and theoretical frameworks, data was collected through
primary to understand their perception towards this financial innovation and to test the influence of
socio-demographic factors on its usage.
A factor analysis revealed that perceptions are shaped on 4 underlying dimensions, namely, Convenience,
Perceived Ease of Use and Perceived Usefulness, being the most prominent, followed by ‘Priva cy, Security
and Lifestyle, ‘Perceived risk and Trust’ and ‘Traditional Banking and Financial Costs’ which was ranked last.
In addition, a binary logistic regression was performed to test the dynamics of socio -economic variables on
mobile banking usage and the best fitted model indicated that marital status and socioeconomic group are
powerful predictors of mobile banking usage.
In addition, it was necessary to understand the basis on which mobile banking users differ from non -users
and a binary logistic regression was performed taking into account their respective socio-demographic
characteristics. The overall results of the core model indicate that marital status and occupational group
play a major role in mobile banking usage. For instance, respondents belonging to higher socioeconomic
classes and who are married with no child are more likely to be users of mobile banking. Another
conclusion that could be drawn from the model is that, individuals in the lower echelons of the
socioeconomic ladder are less likely to be users of mobile banking. In the same line of thought, personal
income was found to be significant predictor of mobile banking adoption in Kenya (Kweyu and Ngare, 2013)
and in Senegal (Fall et al., 2015). Moreover, the findings of Bhatt and Bha tt (2016) also revealed that
mobile banking usage were mostly common among the high income earners who were married but was
more pronounced among women.
Keywords: Mobile Banking, Factor Analysis and Binary Logistic Regression
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 8 ~
INTRODUCTION
With the proliferation of globalisation and advanced technology, the evolution of
wireless internet and mobile devices have consequently transformed the lifestyle of
consumers. The increased penetration of smart phones in our society demonstrates the
degree to which its usage has become ingrained in modern culture, as a dominant tool
for banking, payments, budgeting and shopping (Cruz, 2010). It is undeniable that
financial technology (FinTech) has emerged as a revolution in the world of finance, thus,
altering the way business is conducted to better servicing clients in the new era of digital
financial services. FinTech companies acknowledge the paradigm shift experienced by
the banking sector as millennials are increasingly migrating to digital platforms to carry
out their banking activities (Meola, 2017). As a result, banks have been compelled to
rethink their strategies and tailor their services to meet consumer demands with mobile
friendly financial services. With mobile banking, customers are able to undertake a range
of banking transactions; from checking their accounts, transferring money to paying bills
and managing their finance without visiting physical branches (Sulaiman et. al., 2007).
Moreover, mobile banking is identified is a key tool in contributing to financial inclusion
and economic development in many countries (Daniel, 2015; Siddik et al, 2014).
Worldwide, banks are prioritising financial innovation in order to achieve higher financial
performance via customised banking services (Crabbe et al., 2009).
Aims and Objectives
The principle aim of this study is to gain an understanding of factors influencing the
adoption of mobile banking among bank customers in Mauritius. In order to achieve this
aim, a series of sub-objectives were developed as follows: a) To gauge the mobile
banking landscape of Mauritius i.e., to understand who are the users and non-users of
mobile banking services and identify the barriers and motivators of its adoption or usage.
b) Investigate the factors shaping bank users’ perception towards mobile banking and
contributing to their favourable or unfavourable attitudes. c) Test the influence of
socioeconomic variables in mobile banking usage and develop a model to forecast usage
based on the demographics of respondents.
LITERATURE REVIEW
The definition of mobile banking differs from one researcher to the other. Crabbe et.al.
(2009) defines mobile banking or m-banking as “the ability to perform banking
transactions online on portable mobile devices via Short Messaging Services (SMS) or
Wireless Application Protocol (WAP)”. A report published by the Federal Reserve System
(2016) states that mobile banking encompasses services enabling consumers to obtain
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 9 ~
account information and conduct transactions with their financial institutions, allowing
customers to effect payments for goods and services or even transfer money through
their phone. Furthermore, in a Guideline on Mobile Banking and Mobile Payments
issued by the Bank of Mauritius (2013), mobile banking is outlined as “a bank or mobile
network operator which is authorised to provide services that enable the process of
money transfer and exchange of money for goods and services between two parties
using a mobile communication device”. Therefore, it can be understood that this new
innovative service gives bank customers access to their current and/or savings account
through their mobile phone, offering round-the-clock flexibility and convenience.
Empirical Reviews of Studies Relating to Mobile Banking Usage
In order to understand the drivers of mobile banking usage or adoption, a number of
journal articles were consulted. As such, empirical reviews suggest that there are several
factors accountable for shaping perception, attitudes and behaviours towards the mobile
banking platform. This section provides an overview of those factors influencing mobile
banking adoption.
Application of Conceptual Frameworks
Review of literature on mobile banking suggest that several intention-based models and
theories have been developed to study factors influencing mobile banking adoption
(Rezaei et al., 2013; Jeong and Yoon, 2013; Aboelmaged and Gebba, 2013). The theories
commonly rely on the theory of reasoned action, the theory of planned behaviour, the
technology acceptance model, and the diffused theory of innovation. A brief overview is
given on each theory to serve as construct and to provide illumination on how bank
customers’ perception are shaped and how these factors influence mobile banking
usage.
Theory of Reasoned Action
The Theory of Reasoned Action (TRA) is renowned theory, developed by Ajzen and
Fishbein (1980), which advocates that user acceptance or rejection of an innovation or
system is influenced by attitudes, subjective norms and behavioural intention.
Subjective Norms
Subjective norms englobe perceptions that are formed based on word-of-mouth and can
be expressed as social influences on behaviour of an individual. Subjective norms were
identified as powerful antecedents measuring social influences on behavioural intention
to adopt mobile banking (Balabanoff, 2014). Also, Agarwal and Prasal (1998) argued that
perception towards an innovation and its acceptance are influenced by the individuals’
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 10 ~
level of innovativeness. Thus, mobile banking as financial innovation is likely to be
accepted by individuals who have a positive perception of it.
Attitudes
According to Balabanoff, (2014), attitudes are shaped based on one’s perception of the
system’s usefulness pertaining to relative advantages, risk and privacy and preferred
features. In other terms, attitude is referred to the extent to which one has a favourable
disposition, resulting in an actual open behaviour. Shaikh and Karjaluoto (2015) who
studied influential variables on mobile banking adoption revealed attitude among the
most significant motivators in developed and less developed nations.
Technology Acceptance Model
The Technology Acceptance Model (TAM) was initially proposed by Davis (1985), with
the aim of studying the acceptance of a new technology by an individual. The theoretical
model postulates that users’ acceptance or adoption of the new technology is based on
their “perceived usefulness” and “perceived ease of use” of the technology.
Perceived Usefulness
Perceived usefulness is described as the extent to which individuals believe that using a
particular technology would improve the efficacy of their job performance (Davis, 1985).
Several studies have demonstrated that mobile banking adoption is positively correlated
with perceived usefulness (Siddik et al, 2014, Jeong and Yoon, 2012). The perceived
usefulness relates to the belief on how mobile banking could be beneficial to the user.
Benefits may pertain to saving time as well as carrying banking transactions effectively.
In another study, mobile banking adoption was mediated by the speed of banking
transactions and associated low financial costs (Yang, 2009 cited in Govender and Sihlali,
2014).
Perceived Ease of Use
Conversely, another element which plays a role in mobile banking adoption is perceived
ease of use. This refers to the extent to which an individual is of the opinion that the
technology does not necessitate much physical or mental effort (Davis, 1985). This
model has been adapted by many researchers to understand the dimensions involved in
user adoption behaviour. As such, an application of the TAM in the context of developed
and developing countries uncovered that the main drivers of intention to adopt mobile
banking are “perceived usefulness”, “attitudes” and recently “compatibility” of the
device with the individual’s lifestyle (Shaikh and Karjaluoto, 2015). Another study
conducted by Aboelmaged and Gebba (2013) also revealed that perceived usefulness
significantly influenced mobile banking adoption in Dubai. Chitungo and Munongo (2013)
applied an extended version of the TAM to investigate mobile banking adoption. The
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 11 ~
significant factors were found to be perceived ease of use, relative advantages, personal
innovativeness and social influences.
Theory of Planned Behaviour
The Theory of Planned Behaviour (TBP) was also developed as an extension to the TRA
(Aboelmaged and Gebba, 2013). The TBP is conceived as a model explaining that
behaviour or intention to adopt a particular technology is triggered by a favourable
attitude towards that technology while subject norms and behavioural control are also
influential. Khatimah and Halim (2016) also decomposed the elements of the TPB in
Indonesia to explore the usage of e-money, which revealed that social influences may
positively impact intention to use e-money, in line with the findings of Abdulkadir et al.
(2013) who also found social influences as impactful on mobile banking usage, along
with perceived usefulness.
Theory of Diffused Innovation
The financial services sector is highly dependent on technological services. Therefore,
banks attribute importance to technological innovations so as to ensure their survival
against competition. “The innovation of delivering financial services through mobile
devices represents a complex interaction between an intangible service and technology
based service delivery”, (Black et al., 2001). An exploration of the literature also suggests
that the TDI, developed by Rogers (1995) is widely used to explain mobile banking
adoption. To the author, the process of favouring an innovation is determined by the way
this innovation is communicated, by means of different sources over time or among
members in a social circle. In other words, the proliferation of innovation, which is the
adoption rate is influenced by a set of five factors, namely, relative advantage,
compatibility, complexity, number of trials and observability of the innovation.
Perceived Benefits and Limitations of Mobile Banking
The theories expounded above offer explanations on the process by which attitudes and
perceptions are shaped and how behavioural intention is determined towards mobile
banking. Several themes relating to perceived benefits were encountered as contributors
to mobile banking adoption.
Access and Convenience
Mobile banking is often praised for its ubiquitous bank access that surpasses
geographical limitations (Laforet and Li, 2005; Laukkanen, 2010). Cited as the main factor
driving the adoption of mobile banking, convenience is the most salient advantage that
mobile banking offers to customers while providing them with unrestricted access to
banking facilities (Chitungo and Muningo, 2013). Another study argues that mobile
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 12 ~
banking as a means of branchless banking compensates for the lack of banking
infrastructure in less developed countries (Dermish et al, 2011).
Reduced Transaction Costs
Another advantage that mobile banking provides to customers is lower transactions
costs,
(Polatogulu and Ekin, 2004). It is argued that by bypassing branch channels, the
workload
of branches are minimised, leading to an overall reduction in transaction costs. Thus,
Kabir (2013) tested the concept of comparative advantage in the form of cost and time,
more precisely airtime and bank charges and it was found to be a significant indicator of
mobile banking usage.
Compatibility with Lifestyle
A pertinent factor found to propel adoption rate of mobile banking throughout literature
has been its compatibility with the lifestyle and needs of the user (Shaikh and Karjaluoto,
2015; Al-Jabri and Sohail, 2012). Studies suggest that potential users of mobile banking
are likely to be influenced by their technology driven lifestyle. In the same line of
thought, it is argued that when a developed technology is compatible with lifestyle of
people, the more likely they are to adopt it (Lee and Lee, 2010, cited in Ramdhony and
Munien, 2013). The literature also provided evidence of certain limitations, influencing
one’s perception and attitude unfavourably and hence hindering its adoption.
Some of the limitations are discussed below.
Complexity
Complexity is another component which is similar to the perceived ease of use in the
TAM. Complexity is explained as the perceived level of difficulty to use and understand
the innovation (Gerrard and Cunningham, 2003). Polatoglu and Ekin (2004) discuss the
level of education and perceived complexity of an innovation as being inversely
correlated.
Perceived Risk
Perceived risk has emerged consistently in theory of diffused innovation as a core
construct that influences adoption of mobile banking (Özdemir, 2014; Malhotra, 2011).
However the element of risk has been conceptualised on different axes such as financial
risk, security or privacy and social and time risk in the context of internet banking and
was identified as a main barrier, impeding online banking adoption (Lee, 2008).
Govender and Sihlali, (2014) also discuss risk and privacy as essential players in
obstructing the acceptance of mobile banking.
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 13 ~
Perceived Credibility
The two paramount elements contributing to perceived credibility are security and
privacy (Wang et al, 2003, cited in Abdulkadir, 2013). Perceived credibility refers to the
extent to which an individual believes that using a particular system or technology is not
likely to incur security or privacy breach. While security encompasses the protection of
the system from unauthorised access, privacy refers to the safeguarding of users’
confidential information while using mobile banking.
Trust
Yao et al. (2013) debate on the element of trust and distrust arising with the acceptance
of mobile banking. Distrust was discussed as a potential element deterring mobile
banking adoption. This theme also overlapped with handling of risk and the quality of
the system, playing an influential role in shaping users’ attitudes and perceptions.
Another study (Masrek et al., 2013) purports that satisfaction with mobile banking and
trust in technology are positively related. Moreover, Govender and Sihlali (2014) also
identified trust which could be intertwined with fraud risk, hacking and privacy concerns.
Revelatory Studies relating to Mobile Banking Usage or Adoption
This section provides a review on studies conducted to explore possible behavioural
attitudes and perception towards mobile banking and its usage in different countries.
“Factor Analysis of Customers Perception of Mobile Banking Services in
Kenya”, (Kweyu and Ngare, 2013)
Inspiration was drawn from this study carried out in a Kenyan context, whereby the
mobile banking service, known as “m-Shwari”, is extended to cater for savings, earning
interest and borrowing over short lapse of time through mobile phones. The study
aimed at investigating the factors that influence mobile banking. The authors have used
the constructs of the TAM, along with Roger’s model of diffusion of innovation to
determine consumer’s adoption of this financial innovation. This article discussed the
sub Saharan African region as a potential and niche market for financial development
with mobile banking use, also in line with arguments proposed by Bhatt and Bhatt (2016)
and Medhi et al. (2009). The analysis was performed in different stages whereby an
exploratory factor analysis was firstly used to reveal important factors playing a role in
mobile banking adoption. The rationale behind using factor analysis to the researcher is
to get a better understanding on how many factors are present in the set of variables.
This statistical technique retained seven clusters in total, accounting for Ease of Use,
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 14 ~
Bank product features, Relative advantages, Usefulness, Risk, Interest, Concerns and
Doubts. Another study (Govender and Sihlali, 2014) also found perceived ease of use as
a positive influence to use mobile banking. Furthermore, the findings of the factor
analysis were used and tested for significant differences between genders using
Mann-Whitney test. It was unearthed that mobile banking adoption perceived ease of
use and risk of use were not different between genders. On the other hand, factors like
age group, personal income and educational level are likely to influence usage of mobile
banking.
“Analysing the Mobile Banking Adoption Process among Low-Income
Populations: A Sequential Logit Model”, (Fall et al., 2015)
The primary aim of this study was to shed light on the socio-economic variables that
could provide a justification for mobile banking adoption. The sample selection involved
a two-stage process, whereby the population of interest were mainly residents of the
suburbs of Dakar (Senegal) and included 127 households and 648 individuals. The
sample was spread by age, gender, employment and sector of employment. This study
exploits the different stages that an individual goes through in decision making with
regards to mobile banking. Out of the 648 individuals, 620 were aware of mobile banking,
109 had the service but only 73 were found to be users of the service and a decision tree
model was used to depict this behaviour. Thus, the logit model was applied in 3 phases,
mainly; Awareness, Ownership and Adoption (Usage). In the first stage revealed that
elder generations (above 45 year) were more likely to be unaware of mobile banking
(tested at 10% significance level). In the second phase, high literacy rate, wages, having a
credit and saving scheme and having a micro-enterprise were found to be significant
determinants in ownership of mobile banking. Last stage results confirmed that
education level and income were significant cognitive motivators of mobile banking
adoption. A noteworthy observation was that income did not emerge as an explanatory
variable in acquiring the service, but was found to be among the major determinants
while adopting mobile banking. Education and income were also witnessed to be
associated to mobile banking usage in other contexts (Kweyu & Ngare, 2013; Özdemir,
2014).
“Factors Affecting Malaysian Mobile Banking Adoption: An Empirical
Analysis”, (Cheah et al., 2011)
This empirical study adapts the constructs of the TAM to explore factors triggering
mobile banking adoption in Malaysia. Data was collected through primary sources
among 400 respondents. However, it recorded as valid response of rate of only 44%.
Data was analysed through statistical tools such as multiple regression and factor
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 15 ~
analysis. The factors which were positively correlated to mobile banking adopt relate to
Perceived Usefulness, Relative Advantages, Perceived Ease of Use and Personal
Innovativeness. In the study, mobile banking penetration was estimated at 14.3% only
and Perceived risk was identified as a major obstructer to mobile banking usage. On the
other hand, social norms was found as exhibiting no influence on mobile banking
adoption, constrasting with the revelations of Siddik et al. (2014).
METHODOLOGY
This section will demonstrate the strategies that were utilized to conduct the research
such as the research design, sample size and data collection. To gather primary data a
questionnaire was designed that comprised of four sections: Lifestyle and Financial
Behaviour, Mobile Banking Usage, Attitudes and Perception towards Mobile Banking and
the demographic and was eventually distributed to users and non- users of mobile
banking through online surveys. A sample size of 300 respondent were taken into
consideration. Factor analysis and a logistic regression was used for this research.
ANALYSIS
Factor Analysis
Since it was of interest to uncover the underlying patterns that contribute to the
adoption or usage of mobile banking, an attribute question was asked, relating to the
perceived benefits and shortcomings of mobile banking. Therefore, the factor analysis
was conducted using the variables shown in the table below.
Variable Code Mobile Banking Statements
C1 Can check account balance and perform transactions any time
C2 Can save time and effort of visiting branches
C3 Can perform daily transactions more easily
C4 Mobile banking is a financial innovation
C5 Mobile banking is very easy to learn and navigate
C6 Mobile banking is very user-friendly
C7 Mobile banking is useful to conduct banking
C8 Banking transactions are conducted more quickly
C9 Mobile banking offers more privacy
C10 Mobile banking makes banking faster and safe
C11 Mobile banking always provides accurate information
C12 Mobile banking charges lower transaction fees
C13 Mobile banking fits my lifestyle
C14 Mobile banking increases self-prestige
C15 Mobile banking is subject to hacking and spams
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 16 ~
C16 Mobile banking is financial risky - can lose money
C17 Cannot trust mobile banking services
C18 Traditional banking is as important as mobile banking
C19 It is important to interact with bank staff
C20 Too much costs involved with mobile banking
The scale for the negative questions, such as C15 to C20 were inverted to perform the
analysis.
Measure of Sampling Adequacy
The output of formal statistical tests was studied to evaluate the sampling adequacy and
to verify whether the data is cohesive to the assumptions of the model. Table 1 shows a
value of 0.912 given by the Kaiser-Meyer-Olkin (KMO), exceeding the benchmark of 0.5
and therefore suggests that sufficient correlation exists among the variables.
Table 11: KMO and Barlett's Test Table
Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .912
Bartlett's Test of Sphericity
Approx. Chi-Square 4353.022
df 190
Sig. .000
The Barlett’s test of Sphericity displays a p-value of 0.000, highly significant to reject the
hypothesis that the correlation matrix is an identity matrix. Thus, these measures
confirm that the factor analysis is a robust model to study mobile banking perceptual
dimensions.
Factor Retention
The Kaiser’s rule recommends to retain factors with Eigen values exceeding 1.The first
factor retained accounts for the maximum variation (30%) in the data while the
subsequent factors account for lesser variation. However, the second factor explains an
additional 18% of variation and cumulatively accounts for 48% of variation in total.
However, each factor captures the maximum variance but remains uncorrelated with
other factors. It is observed that 4 variables were retained (Eigen value > 1), accounting
for 72% of total variation in the data.
Factor Rotation
It was noted that most variables’ loadings were spread among the first to the third factor,
while the fourth factor did not inherit any loading. Therefore, the Varimax rotation was
applied. This factor rotation technique maximises the variation captured by each factor
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 17 ~
and redistributes the total amount of variance over the newly extracted four factors.
Following this re-orientation, the Rotated Factor Matrix was produced where the
structure of the loadings matrix is simplified. Each variable was loaded on their
corresponding factor for ease of interpretation as clusters of variables.
The Model
The table below displays the overall factor analysis results.
It can be noted that the first factor, capturing a larger share of variance, has clustered
variables relating to ‘Convenience’, ‘Perceived Ease and Use’ and ‘Perceived Usefulness’.
Therefore, these elements are the most influential in shaping people’s perception
towards mobile banking. The second factor, accounting for less variation has been
labelled as ‘Privacy’, ‘Security’ and ‘Lifestyle’ and shows that the privacy and security
that mobile banking offers correspond to people’s lifestyles. The third factor retained in
the model groups variables relating to ‘Perceived Risk’ and ‘Trust’ while the fourth factor,
explaining the least variation relates to aspects of ‘Traditional Banking’ and ‘Financial
•C1 - Can check account balance and perform transactions any time
•C2 - Can save time and effort of visiting branches •C3 - Can perform daily transactions more easily
•C4 - Mobile banking is a financial innovation
•C5 - Mobile banking is very easy to learn and navigate
•C6 - Mobile banking is very user-friendly •C7 - Mobile banking is useful to conduct banking
•C8 - Banking transactions are conducted more quickly
Factor 1: Convenience, Perceived Ease of Use and Perceived Usefulness
•C9 - Mobile banking offers more privacy
•C10 - Mobile banking makes banking faster and safe •C11 - Mobile banking always provides accurate information
•C12 - Mobile banking charges lower transaction fees
•C13 - Mobile banking fits my lifestyle
•C14 - Mobile banking increases self-prestige
Factor 2: Privacy, Security and Lifestyle
•C15 - Mobile banking is subject to hacking and spams
•C16 - Mobile banking is financial risky - can lose money •C17 - Cannot trust mobile banking services
Factor 3: Perceived Risk and Trust
•C18 - Traditional banking is as important as mobile banking
•C19 - It is important to interact with bank staff •C20 - Too much costs involved with mobile banking
Factor 4: Traditional Banking and Financial Cost
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 18 ~
costs’.
Application of the Binary Logistic Regression
The core model which might enable us to gauge mobile banking adoption is developed
in this sub-section. Inspiration was drawn from Fall et al. (2015) and Khan et al. (2017) to
study the influence of socio-demographic variables on mobile banking adoption. The
response variable being ‘Usage’ is based on whether an individual uses mobile banking
or not. Since it accommodates binary responses (Yes/ No), it is considered to be
bounded and therefore employing a generalised linear model for the purpose of analysis
was suitable. As discussed in the previous chapter, such unified framework caters for
studies of this nature with the binary logistic model, which is reliable in forecasting
dichotomous outcomes. The explanatory variables postulated to be related to mobile
banking adoption or usage (response variable) were coded as displayed in Table 2 for the
analysis.
Table 2: Response variable and explanatory variables
Variables Details Code
Use mobile banking Yes 1
No 0
Gender Male 1
Female 2
Area of residence Urban 1
Rural 2
Age Generation Z ( 18-20) 1
Generation Y (21-34) 2
Generation Z (35-49) 3
Baby boomers (50-64) 4
Head of Household Yes 1
No 0
Ethnic group Hindu 1
Muslim 2
General population 3
Chinese 4
Socio Economic Group AB - Upper class 1
C1 - Upper middle class 2
C2 - Lower middle class 3
DE - Lower class 4
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 19 ~
Marital status Single 1
In a relationship 2
Married with children 3
Married without children 4
Divorced/ separated/ widowed 5
Education level No formal education 1
Primary 2
Secondary 3
Tertiary 4
Reduction in Deviance Test
For the purpose of this analysis, the statistical software R was used, whereby all the
logistic regression were run. Table 3 summarises the results from the series of
regressions. The motive of this first step analysis was to uncover the variables that had
significant reduction in deviance, measured by the G-statistics (with a Chi-square
distribution), along with significant p-values (p<0.01).
Table 3: Reduction in Deviance
Explanatory Variable G-statistic p-value
Gender 0.32063 0.5712
Area of residence 32.692 0.000
Age 107.66 0.000
Education level 67.264 0.000
Ethnic group 53.622 0.000
Marital status 17.899 0.001
Head of household 15.812 0.000
Socio-economic group 104.61 0.000
Surprisingly, most socio-demographic variables were found to be statistically significant
(p < 0.01) with the exception of gender. Therefore, mobile banking usage could be
explained by an individual’s area of residence, age, level of education, ethnicity, marital
status, his/her socio-economic group and if he or she is the head of household. Since
gender had poor explanatory power (low reduction in deviance and high p-value), it was
withdrawn from the model. However, age and socio-economic group were observed to
have higher reduction in deviance and therefore, good explanatory power.
Multicollinearity Tests
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 20 ~
Moreover, as a standard measure, it was important to test for multi-collinearity among
variables so as to avoid spurious interpretation of possible relationships between mobile
banking usage and socio-demographic variables. To test the relationship among nominal
and ordinal variables, the Pearson-Chi Square was used. The correlation coefficient
depicts the strength of a linear positive or negative relationship between the variables.
Correlation coefficients nearing zero have a small or moderate correlation and its
significance value (p-value) is compared to 0.01, whereby if p-value is greater than 0.01,
it implies that there is enough evidence that correlation observed is not statistically
significant. It was noted that area of residence was mildly correlated with ethnicity,
marital status and head of household but on the other hand, ethnic group and marital
status were found to be correlated with head of household at 1% significance level.
Thus, these variables could not be considered in conjunction. However, moderate
correlations were observed for marital status with education level and marital status
with socio-economic group.
Ordinal variables were treated separately to study any pairwise associations using the
Spearman Rank correlation coefficient. It was found that a high correlation exists among
all of the variables (low p-values (<0.01), indicating that all of the ordinal variables could
not be included in the model at the same time. Associations between variables such as
age and education level and socio-economic group were found to be highly significant
(at 1% level). Moreover, the variable age was also found to be associated with the
nominal variables and therefore was to be eliminated from the model.
Choice of Model
After eliminating the correlated variables, the following model combinations were
proposed:
Mobile Banking Usage = f (Residence, Ethnic Group)
Mobile Banking Usage = f (Residence, Marital status)
Mobile Banking Usage = f (Residence, Head of Household)
Mobile Banking Usage = f (Education, Marital Status)
Mobile Banking Usage = f (Socio-Economic Group, Head of Household)
Mobile Banking Usage = f (Socio-Economic Group, Marital Status)
However, after taking into consideration variables which have maximum reduction in
deviance, it was desirable to consider socio-economic group and marital status in the
model, where by,
Mobile Banking Usage = f (Socio-Economic Group, Marital Status)
Logit
= α+ , where α is a constant
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 21 ~
The Null Model
The first step to constructing the model was to start with a neutral model, which is
known as the Null model, having maximum deviance and which is given by an intercept
only model:
Logit
= α
The Null model postulates that explanatory variables have zero correlation coefficients.
In other words, the likelihood of adopting mobile banking would be equally spread
among all socio-demographic characteristics. The null model displays a maximum
deviance of 42.869 and the objective would be to study the deviations from this
neutrality while adding variables to this model. To understand the behaviour of the
model, the following formula was used to calculate the likelihoods.
Where the logit is given by 0.2796 and the likelihood is calculated as
= 0.5694,
implying that the likelihood of adopting mobile banking, irrespective of marital status
and socio-economic group is 57%.
One Factor Model
The response variable ‘Usage’ was regressed in turn on each of the considered variable
and the different models are elaborated below.
Mobile Banking Usage versus Marital Status
The variable ‘Marital Status’ consisted of 5 modalities, namely; Single, In a Relationship,
Married with Children, Married without Children and Widowed/ Divorced/ Separated.
The objective was to enquire who would be more inclined to use mobile banking. The
first model is given as follows:
a. Mobile Banking Usage = constant + Marital Status
Logit
= α + , where i = 1, 2, 3, 4, 5
The model posits that usage of mobile banking is a function of marital status only and
the table below shows the output results from the binary logistic model.
Categories Logits ( Probabilities
Single 0.4383
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 22 ~
In a relationship 0.4383 – 0.2664 =
0.1719
Married with
children 0.4383 – 0.3512 = 0.0871
Married with no
child 0.4383 + 0.1343 = 0.5726
Divorced/
widowed/
separated
0.4383 – 0.7006 = –
0.2623
For instance, the likelihood of using mobile banking if individual is single is 0.6079 and
the likelihood of using mobile banking if individual is married with no child is 0.6394.
From the above table, we observe that > , which means
there is a stronger likelihood that an individual adopts or uses mobile banking if he or
she married with no child. The probability of mobile banking adoption among single
individuals is also quite close. Alternatively, those who are divorced, widowed or
separated are less likely to use this platform.
G-statistics Implication
The hypotheses tested under the first model is:
: There is no relationship between mobile banking usage and marital status
: There is a relationship between mobile banking usage and marital status
The critical value for at 4 degrees of freedom and 1% level of significance is at
13.277. The G-statistic, reveals a reduction in deviance of 3.749 which is less than the
critical value and therefore insignificant. Therefore, it suggests that there is not enough
evidence to reject and conclude that there is a relationship between mobile
banking usage and marital status. Therefore, this leads us to test the relationship with
the next variable under the one factor model.
Mobile Banking Usage and Socio-Economic Group
The next step was to regress ‘Usage’ with the second variable of interest which is
‘Socio-Economic Group’. This explanatory variable accommodated 4 responses, mainly
the AB, C1, C2 and DE. It was of interest to enquire in which socio-economic group,
mobile banking usage is more prominent. Therefore, the second model was given as
follows:
b. Mobile Banking Usage = constant + Socio-Economic Group
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 23 ~
Logit
= α + , where j = 1, 2, 3, 4
The model postulates that mobile banking usage is a function of socio-economic group
only. Hence, the output from the binary logistic results is calculated as shown below.
Categories Logits ( Probabilities
AB – Upper Class 1.6650
C1 – Upper Middle
Class
1.6650 – 1.3337 = 0.3313
C2 – Lower Middle
Class
1.6650 – 1.7295 = –
0.0645
DE – Lower Class 1.6650 – 2.0899 = –
0.4249
From Table 19, it can be noted that the probability of an individual using mobile banking
if the latter belongs to the AB class is 0.8409. Also, the table shows that
, suggesting that as we go down the socio-economic ladder, an individual is
less likely to use mobile banking. Those belonging from lower working class (DE) were
identified to be less likely to use mobile banking, with a usage probability of 0.3953.
G-statistics Implication
The hypotheses tested under the second model is:
: There is no relationship between mobile banking usage and socio-economic group
: There is a relationship between mobile banking usage and socio-economic group
The critical value for at 3 degrees of freedom and 1% level of significance is at
11.345. Since the G-statistic or reduction in deviance of 22.035 exceeds the critical
value, there is therefore enough evidence to reject and conclude that there is an
association between mobile banking usage and socio-economic group. Hence,
socio-economic group was found to be a powerful predictor of mobile banking usage.
Two Factor Model
Following the results obtained from the One Factor models, it was of interest to discover
the dynamics of an additive model, using the variables ‘Marital Status’ and
‘Socio-Economic Group’ altogether. The logit for the two-factor model is given as follows:
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 24 ~
c. Mobile Banking Usage = constant + Marital Status + Socio-Economic Group
Logit
= α +
,
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 25 ~
The table shows the Probabilities Interpretation in Two-Factor model
Characteristics
Marital Status Socio-Economic Group
Single AB – Upper Class
In a relationship AB – Upper Class
Married with children AB – Upper Class
Married with no child AB – Upper Class
Divorced/ widowed/
separated AB – Upper Class
Single C1 – Upper Middle Class
In a relationship C1 – Upper Middle Class
Married with children C1 – Upper Middle Class
Married with no child C1 – Upper Middle Class
Divorced/ widowed/
separated
C1 – Upper Middle Class
Single C2 – Lower Middle Class
In a relationship C2 – Lower Middle Class
Married with children C2 – Lower Middle Class
Married with no child C2 – Lower Middle Class
Divorced/ widowed/
separated
C2 – Lower Middle Class
Single DE – Lower Class
In a relationship DE – Lower Class
Married with children DE – Lower Class
Married with no child DE – Lower Class
Divorced/ widowed/
separated
DE – Lower Class
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 26 ~
Saturated Model
The saturated model incorporates a multiplicative model and postulates that mobile
banking usage is a function of both Marital Status and Socio-Economic Group and their
interaction.
d. Mobile Banking Usage = constant + Marital Status + Socio-Economic Group +
Marital Status*Socio-Economic Group
Logit
= α +
Such model is a perfect fit for the data and has zero deviance (maximum reduction in
deviance), but however the principle of parsimony debates on the acceptable number of
variables that can be fitted into a model.
The table below depicts the deviance of all models, starting with the null model to the
saturated model.
Model Logit ( Residual
Deviance
Degrees of
Freedom
Null model α 42.869 19
One factor
Marital Status (M) α + 39.120 4
Socio-Economic Group (S) α 20.834 3
Two factor
M + S α + 16.121 7
Saturated model
M + S + M*S α + 0 0
Goodness of Fit
Deviance is used as a tool to measure the adequacy of a model. Thus, in order to choose
the model which was best fitted for this study, the following hypotheses were tested to
evaluate each model’s goodness of fit.
: There is no lack of fit
: There is lack of fit
The tests were carried out at 1% significance level and the below table was produced as
a result.
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 27 ~
Goodness of Fit Table
Model Residual
Deviance
Degrees of
Freedom
at 1%
significance level
Degree of fit
Null model 42.869 19 36.191 Lack of fit
One factor
Marital Status (M) 39.120 4 13.277 Lack of fit
Socio-Economic Group (S) 20.834 3 11.345 Lack of fit
Two factor
M + S 16.121 7 18.475 Adequate fit
Saturated model
M + S + M*S 0 0 0 Perfect fit
Where the residual deviance was greater than the value, the null hypothesis was
rejected in favour of the alternative hypothesis. Results obtained for the null model are
conclusive enough to suggest that model is not a good fit as the deviance exceeds the
critical value at 19 degrees of freedom (42.869 > 36.191) at 1% significance level. Thus,
this warns us that the null model is not appropriate for the study and that another
model should be looked for. While introducing the variable ‘Marital Status’ into the
model, the deviance declined to 39.120 on 4 degrees of freedom. However, high
deviance exceeding the value at 1% significance level suggests that there is not
enough evidence to reject in favour of . Therefore, it could be concluded that
‘Marital Status’ on its own in the model was not a good fit. When the variable
‘Socio-Economic Group’ was studied, the deviance reduced to 20.834 on 3 degrees of
freedom. However, this deviance was still found to be insignificant (20.834 >15.086).
Therefore, was rejected as there was enough evidence to conclude that there is a
lack of fit in the model.
Even though ‘Marital Status’ and ‘Socio-Economic Group’ do not perform well in the
model on their own, it was of interest to test them in conjunction to see their dynamics
under a two-factor model. From the above table, we observe that the deviance declined
to 16.121 on 7 degrees of freedom, in which case is smaller than the value (i.e.
16.121 < 18.475). Thus, there was enough evidence to reject and conclude that
the two factor model is an adequate fit.
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 28 ~
Since ‘Marital Status’ and ‘Socio-Economic Group’ increase the precision of the model
altogether, this was selected as the best model and is presented as follows:
Logit
= α +
whereby,
α : logit for the reference category for Marital Status and ‘Socio-Economic Group’
: change in the logit of the probability associated with one unit change in the
explanatory variable (Marital Status), with other variables constant
: change in the logit of the probability associated with one unit change in the
explanatory variable (Socio-Economic Group), with other variables constant
Therefore, probability table for best model was constructed as shown below from the
equation obtained for the best model which explains ‘Mobile Banking Us age’ in terms of
marital status and socio-economic group and also calculates the probabilities for each
category.
Socio-Economic
Group (j) Marital Status (i)
Coefficients
(ij) Logit Odds Probability
AB
Single 1.8097 6.1086 0.8593
In a relationship -0.4506 1.3591 3.8927 0.7956
Married with
children -0.5445 1.2652 3.5438 0.7799
Married with no
child 0.1583 1.968 7.1563 0.8774
Divorced/
widowed/
separated
-0.6679 1.1418 3.1324 0.7580
C1
Single -1.2488 0.5609 1.7522 0.6367
In a relationship -1.6994 0.1103 1.1166 0.5275
Married with
children -1.7933 0.0164 1.0165 0.5041
Married with no
child -1.0905 0.7192 2.0528 0.6724
Divorced/
widowed/
separated
-1.9167 -0.107 0.8985 0.4733
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 29 ~
C2
Single -1.7835 0.0262 1.0265 0.5065
In a relationship -2.2341 -0.4244 0.6542 0.3955
Married with
children -2.3280 -0.5183 0.5955 0.3732
Married with no
child -1.6252 0.1845 1.2026 0.5460
Divorced/
widowed/
separated
-2.4514 -0.6417 0.5264 0.3449
DE
Single -2.1636 -0.3539 0.7019 0.4124
In a relationship -2.6142 -0.8045 0.4473 0.3091
Married with
children -2.7081 -0.8984 0.4072 0.2894
Married with no
child -2.0053 -0.1956 0.8223 0.4513
Divorced/
widowed/
separated
-2.8315 -1.0218 0.3599 0.2647
The best model explained that those who are married with no child and who are from
the upper social class (AB) are more likely to use mobile banking, as compared to
individuals from lower social class, who are divorced/ widowed/ separated.
CONCLUSION
The study had the aim of shedding light on mobile banking usage and its perception in a
Mauritian context. A survey was carried out on a national level among 300 respondents,
through the combination of an online survey platform and face-to-face interviews. The
results have uncovered 186 mobile banking users, who are millennial residents in the
urban regions, who have completed tertiary education and mostly come from upper
class families. This banking facility is mainly used to verify their account balance, acc ount
details and recharge their mobile phone credit. A daily usage is recorded among users.
On the other hand, non-usage of mobile banking is mainly attributed to ostensible
preference for traditional banking and non-users’ indifference to the service. The
security aspect of mobile banking emerged as a third barrier. Moreover, factor analysis
revealed that respondents’ perceptions towards mobile banking are shaped by a four
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 30 ~
dimension criteria. These factors have been described as ‘Convenience, Perceived Ease
of Use and Perceived Usefulness’, ‘Privacy, Security and Lifestyle’, ‘Perceived Risk and
Trust’ and lastly ‘Traditional Banking and Financial Costs’. In additional, a binary logistic
regression which explored the demographic influences on mobile banking usage showed
that socio-economic group and marital status were significant predictors. The model
developed predicts that people who are married with no child, belonging from the upper
working class are prone to be users of mobile banking. However, it is undeniable that
banks are committed to meeting the needs of customers in this digital era, but in order
to foster a culture of mobile banking, awareness should be increased and its
complexities should be minimised. Banks should be more transparent while keeping the
public informed about securitised exchanges and control measures to ensure the
safeguard of personal financial data. Also, since mobile banking was identified as an
essential ingredient for financial inclusion, banks should devise strategies to reach those
from the lower socio-economic class.
REFERENCES
Abdulkadir, N., Galoji, S. I. & Razak, R. B. A., 2013. An Investigation into the Adoption of
Mobile Banking in Malaysia. American Journal of Economics, 3(3), pp. 153-158.
Aboelmaged, M. G. & Gebba, T. R., 2013. Mobile Banking Adoption: An Examination of
Technology Acceptance Model and Theory of Planned Behavior. International
Journal of Business Research and Development, 2(1), pp. 35-50.
Agarwal, R. & Prasad, J., 1998. A Conceptual and Operational Definition of Personal
Innovativeness in the Domain of Information Technology. Information Systems
Research. 9(2), pp. 204-215.
Ajzen, I. & Fishbein, M., 1980. Understanding attitudes and predicting social behavior.
Englewood Cliffs,: NJ: Prentice-Hall.
Al-Jabri, M. I. & Sohail, S. M., 2012. Mobile banking adoption: application of diffusion of
innovation theory. Journal of Electronic Commerce Research, 13(4), pp. 379-391.
Balabanoff, G. A., 2014. Mobile Banking Applications: Consumer Behaviour, Acceptance
and Adoption Strategies in Johannesburg, South Africa (RSA). Mediterranean
Journal of Social Sciences, 5(27), pp. 247-258.
Bank Of Mauritius, 2013. Guideline on Mobile Banking and Mobile Payment Systems, s.l.:
BANK OF MAURITIUS.
Bhatt, A. & Bhatt, S., 2016. Factors Affecting Customers Adoption of Mobile Banking
Services. The Journal of Internet Banking and Commerce, Volume 21, p. 161.
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 31 ~
Black, N. J., Lockett, A., WINKLHOFER, H. & ENNEW, C., 2001. The adoption of Internet
financial services: a qualitative study. International Journal of Retail&
Distribution, 29(8), pp. 390-398.
Cheah, C. M., TEO, A. C., OON, K. H. & TAN, B. I., 2011. Factors Affecting Malaysian
Mobile Banking Adoption: An Empirical Analysis. International Journal of
Network and Mobile Technologies, 2(3).
Chitungo, S. K. & Munongo, S., 2013. Extending the Technology Acceptance Model to
Mobile Banking Adoption in Rural Zimbabwe. Journal of Business Administration
and Education, 3(1), pp. 51-79.
Chitungo, S. K. & Munongo, S., 2013. Extending the Technology Acceptance Model to
Mobile Banking Adoption in Rural Zimbabwe. Journal of Business Administration
and Education, 3(1), pp. 51-79.
Crabbe, M., Standing, C., Standing, S. & Karjaluoto, H., 2009. An adoption model for
mobile banking in Ghana. Int. J. Mobile Communications, 7(5), pp. 515-543.
Cruz, P., 2010. Mobile banking rollout in emerging markets: evidence from Brazil.
International Journal of Bank Marketing, 28(5), pp. 342-371.
Daniel, N., 2015. Relationship between Mobile Banking and Financial Inclusion in Kenya.
s.l.: Theis (Masters). University of Nairobi, Nairobi.
Davis, F. D., 1985. A Technology Acceptance Model for Empirically Testing New End User
Information Systems: Theory and Results. s.l.: Thesis for Ph.D. Degree.
Massachusetts Institute of Technology.
Dermish, A., Kneiding, C., Leishman, P. & MAS, I., 2011. Branchless and Mobile Banking
Solutions for the Poor: A Survey. Innovations, 6(4), pp. 81-98.
Fall, F., Birba, O. & KY, Y., 2015. Analyzing the Mobile-Banking Adoption Process among
Low-Income Populations: A Sequential Logit Model. Economics Bulletin, 35(4),
pp. 2085-2103.
Govender, I. & SIHLALI, W., 2014. A Study of Mobile Banking Adoption among University
Students Using an Extended TAM. Mediterranean Journal of Social Sciences, 5(7),
pp. 451-459.
Jeong, B.-K. & YOON, T. E., 2013. An Empirical Investigation on Consumer Acceptance of
Mobile Banking Services. Business and Management Research, 2(1), pp. 31-40.
Kabir, M. R., 2013. Factors Influencing the Usage of Mobile Banking: Incident from a
Developing Country. World Review of Business Research, 3(3), pp. 96-114.
Khan, S. N., Akter, M. & Akter, R., 2017. Factors Influencing Adoption and Usage of
Mobile Banking: A Bangladesh Experience. International Journal of Finance and
Banking Research, 3(1), pp. 1-12.
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 32 ~
Khatimah, H. & Halim, F., 2016. The effect of attitude and its decomposed, subjective
norm and it decomposed on intention to use E-money server in Indonesia.
Journal of Scientific Research and Development, 3(1), pp. 21-32.
Kweyu, M. & Ngare, P., 2013. Factor Analysis of Customers Perception of Mobile
Banking Services in Kenya. Journal of Emerging Trends in Economics and
Management Sciences, 5(1), pp. 1-8.
Laforet, S. & LI, X., 2005. Consumer attitudes towards online and mobile banking in
China. International Journal of Bank Marketing, 23(5), pp. 362-380.
Laukkanen, T. & Cruz, P., 2010. What determines mobile banking non-adoption?.
[Online] Available at:
http://www.anzmac2010.org/proceedings/pdf/ANZMAC10Final00387.pdf
Lee, M., 2008. Factors influencing the adoption of internet banking: An integration of
TAM 3 and TPB with perceived risk and perceived benefit. Electronic Commerce
Research and Applications.
Masrek, M. N., Mohamed, I. S. & Daud, N. M., 2013. Technology Trust and Mobile
Banking Satisfaction: A Case of Malaysian Consumers. Malaysia, International
Conference on Innovation, Management and Technology Research.
Medhi, I., Ratan, A. & Toyama, K., 2009. Mobile-Banking Adoption and Usage by
Low-Literate, Low-Income Users in the Developing World. Berlin, Springer.
Meola, A., 2017. Mobile fintech is here to stay with banking and payment services.
[Online] Available at:
http://www.businessinsider.com/mobile-fintech-is-here-to-stay-withbanking-and
-payment-services-2016-12?IR=T
Özdemir, N. Ö., 2014. Adoption of Mobile Banking in Turkey. s.l.:Thesis (Masters).
Bahceshir University, Republic of Turkey.
Polatoglu, V. N. & EKIN, S., 2004. An empirical investigation of the Turkish customers‟
acceptance of internet banking services. International Journal of Bank Marketing,
19(4), pp. 156-165.
Ramdhony, D. & Munien, S., 2013. An Investigation on Mobile Banking Adoption and
Usage: A Case Study of Mauritius. Kuala Lumpur, Malaysia, Proceedings of 3rd
Asia-Pacific Business Research Conference.
Rezaei, H. D. A., Kabiry, N. & Forghani, M. H., 2013. Factors Affecting Isfahanian Mobile
Banking Adoption Based on the Technology Acceptance Model. International
Journal of Academic Research in Business and Social Sciences, 3(5), pp. 611-623.
Rogers, F. J., 1995. Diffusion of Preventive Innovations Addictive Behaviors. Volume 27,
pp. 989-993.
Shaikh, A. & Karjaluoto, H., 2015. Mobile banking adoption: A literature review.
Telematics and Informatics, 32(1), pp. 129-142.
The 5th
IBSM International Conference on Business, Management and Accounting 19-21 April 2018. Hanoi University of Industry, Vietnam.
~ 33 ~
Siddik, M. N. A., Sun, G., Yanjuan, C. & Kabiraj, S., 2014. Financial Inclusion through
Mobile Banking: A Case of Bangladesh. Journal of Applied Finance & Banking,
4(6), p. 109-136.
Sulaiman, A., Jaafar, N. I. & Mohezar, S., 2007. An overview of mobile banking adoption
an overview of mobile banking adoption. Int. J. Mobile Communications, 5(2).
Yao, H., Liu, S. & Yuan, Y., 2013. A Study of User Adoption Factors of Mobile Banking
Services Based on the Trust and Distrust Perspective. International Business and
Management. 6(2), pp. 9-14